Rethinking Contrastive Learning in Graph Anomaly Detection: A Clean-View Perspective

2025-05-23

AI Summary

In graph anomaly detection, spurious edges violate the positive/negative sample assumptions of contrastive learning, degrading representations of normal patterns. To address this, we propose Clean-View Graph Anomaly Detection (CVGAD), the first framework to reconstruct graph contrastive learning from a “clean-view” perspective. CVGAD introduces a multi-scale spurious-source identification module to explicitly model noise origins and a progressive edge purification mechanism that iteratively evaluates edge importance and selectively removes spurious edges—mitigating bias inherent in single-step pruning. Integrating anomaly-aware attention, multi-scale GNNs, and contrastive learning, CVGAD jointly optimizes graph structure and representation learning. Evaluated on five benchmark datasets, CVGAD achieves average AUC improvements of 3.2–7.8 percentage points over state-of-the-art methods, empirically validating that spurious-edge purification is critical for enhancing representation robustness.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Computer Vision: Adversarial Attacks & Robustness
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Web data integration and cleaning
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Graph anomaly detection aims to identify unusual patterns in graph-based data, with wide applications in fields such as web security and financial fraud detection. Existing methods typically rely on contrastive learning, assuming that a lower similarity between a node and its local subgraph indicates abnormality. However, these approaches overlook a crucial limitation: the presence of interfering edges invalidates this assumption, since it introduces disruptive noise that compromises the contrastive learning process. Consequently, this limitation impairs the ability to effectively learn meaningful representations of normal patterns, leading to suboptimal detection performance. To address this issue, we propose a Clean-View Enhanced Graph Anomaly Detection framework (CVGAD), which includes a multi-scale anomaly awareness module to identify key sources of interference in the contrastive learning process. Moreover, to mitigate bias from the one-step edge removal process, we introduce a novel progressive purification module. This module incrementally refines the graph by iteratively identifying and removing interfering edges, thereby enhancing model performance. Extensive experiments on five benchmark datasets validate the effectiveness of our approach.

Problem

• Addressing noise from interfering edges in graph anomaly detection
• Improving contrastive learning by identifying interference sources
• Enhancing detection via progressive edge removal purification

Innovation

• Clean-View Enhanced Graph Anomaly Detection framework
• Multi-scale anomaly awareness module
• Progressive purification module for edge refinement